1. Introduction: What precaution, and for what?

This report discusses the relevance and applications of the precautionary principle (PP) in the context of the energy transition, i.e. the move to low-carbon sources and carriers of energy, and the deeper integration of carbon abatement initiatives that compensate for the CO2 already present or being emitted.

In the context of increasingly severe climate change and accelerating global warming, with very dangerous feedback-loop effects that further aggravate the situation, facilitating the transition to “net zero” (or even net negative emissions that reduce atmospheric CO2 levels) is a vital priority. However, it is one that is proving difficult to achieve; the replacement of current technologies by other, low-carbon ones, is hampered by many economic, social, political, and technical obstacles. For many of these challenges, the questions around risk, uncertainty, and precaution play an important role.

The precautionary principle (and precautionary approaches, more generally)1 has generated significant controversy over the years, and the debate around its relevance and effectiveness remains polarised. This report does not intend to argue for or against the PP; instead, it acknowledges, as a starting point, that precautionary approaches are part of the options available to governments and regulators seeking to achieve public policy goals such as energy transition.

The report examines a selection of relevant research, empirical evidence, and existing practice with regard to the use of the PP in a range of policy areas including energy. It does so with a view to enabling a constructive and appropriate application of the principle. In this context, the report focuses particularly on contributing to an improved understanding of the conditions under which the PP can constitute an effective decision-making tool, including the political choices that are inherent to its use. The report focuses particularly on the need for approaching precaution holistically. This approach notably entails considering risk-risk trade-offs and reconciling the need to guarantee a sufficient level of safety with the imperative to capitalise on innovation that addresses climate risks and the irreversible damage it can cause.

As such, the report does not attempt to provide a comprehensive account of all relevant research or practical work pertaining to the PP. Instead, it highlights selected studies as well as concrete examples, draws a number of conclusions, and formulates guidance with a view to informing public policy decision-making.

The report’s specific focus is on the energy transition and, more broadly, climate change mitigation, adaptation and resilience. Thus, while it looks at the construction, interpretation and application of the PP across a range of policy areas, it specifically tries to draw lessons that are relevant to energy and climate. Indeed, among the many difficulties slowing the roll-out of technology that could reduce current or future CO2 levels and potentially limit the increase in global temperatures, the question of precaution is important. Complexity related to applying the PP for the energy transition is particularly high given that the PP is typically used as a tool designed to assess a technology, sector or application. In contrast, climate change is a vastly broader phenomenon that can be considered the sum of global modernisation. As such, no single regulator may have a clear view of related risks. Many technologies are seen, rightly or wrongly, as too hazardous by large swathes of the public and decision-makers alike. A number of these technologies are also, again rightly or wrongly, criticised for having potential negative consequences that are too hard to predict. The report thus seeks to investigate and discuss which could be the “proper” applications of the PP, attempting to distinguish more clearly these from cases where risk-management is more relevant than “precaution” as such. Moreover, it attempts to help think through the issue of balancing risks and uncertainties that exist in both directions, i.e. acting / using the technology versus doing nothing / not using it. Indeed, one of the most important characteristics and challenges of the climate emergency is that “doing nothing” is often not the most “precautionary” approach.

The literature review and analysis presented in this report point to a number of conclusions and highlight the need to further help governments and regulators apply the PP constructively and appropriately — particularly with regards to guidance on the use of precaution in the specific context of the energy transition.

If the PP is to constitute a useful and practical tool for decision-making, its application needs to be guided by the identification of serious potential hazard. This, in turn, requires enhanced systemic readiness (including anticipatory approaches), adaptive learning, periodic assessment (built on an appropriate understanding of the state, pressures, and trajectory of the context), and review.

In addition, adopting holistic and transparent approaches to risk and precaution should be an overarching priority for policymakers and regulators. In this context, “holistic approaches” comprise two essential components. The first is a full and systematic consideration of risk-risk trade-offs, as well as the need to shift focus and available resources — to the extent possible — to actual regulatory and impact outcomes (as opposed to an exclusively “implementing the law” focus). The second essential component involves acquiring a sound understanding of the socio-political context surrounding risk- and precaution-related decision-making. This, in turn, requires extensive public deliberation and stakeholder engagement along the lines of the “concern assessment” notion.

This section puts forward a set of suggestions pertaining to the above-mentioned analysis and priorities. It includes consideration of their institutional implications.

The efforts of governments and regulators should focus -— as a priority — on risk-risk trade-offs, in addition to systemic and cumulative risks. This contrasts to a traditional focus on “individual silos of risk” (Baldwin, 2016[1]). Doing so will lead to better regulatory decisions that help reduce multiple risks in concert (Wiener, 2020[2]).

This refocusing notably involves ensuring that as systemic and informed-but-objective a view as possible is taken of the context in which precaution may be needed — something which requires careful and conscious multi-dimensional leadership overview. Given the resource implications (e.g. capacity building) of developing and implementing more holistic and inclusive approaches, effort should be concentrated on those technologies or innovations with the broadest scope of application and highest potential impacts.

A more holistic approach also involves considering the trade-off between precaution’s potential negative consequence related to stifling innovation, and the need to actively encourage innovation that can deliver the solutions needed to help address major societal challenges such as climate change and the energy transition. Evidence suggests that, if applied overly rigidly, i.e. when an overly-cautious approach is taken without sufficient consideration for risk-risk trade-offs and potentially less restrictive regulatory solutions, the PP can thwart potentially beneficial innovation (Institute for Safety, 2021[3]) (Hydrogen Safety Innovation Programme, 2020[4]).

In addition, there is a strong need to make the tensions and contradictions between different goals and aspirations clear and visible (Blanc, F. et al., 2015[5]). Doing so involves targeting the factors driving the tendency by regulators “to focus on one risk at a time and neglect side effects” — despite existing methodological guidance for more holistic approaches, e.g. (Graham, J. and Wiener, J., 1995[6]). Drivers identified in the literature notably include: “mission-driven agencies, sometimes with narrow legal authority; fragmented institutions, with separate specialised domains; narrow or bounded thinking, driven in part by heuristic errors and in part by decision costs; and the omitted voices of those affected” (Graham, J. and Wiener, J., 1995[6]).

Additional analysis providing a deeper understanding of the reasons behind the limited effective application of available guidance and tools (as well as the potential steps that could be taken to address them), would therefore be valuable — notably with regards to the role of pre-existing structures, practices and philosophies of government, as well as other institutions such as insurance bodies. In addition, ensuring more holistic approaches to risk and precaution will likely require adapting the mandates of the institutions concerned. This could include introducing explicit requirements, so risk-risk trade-offs are better taken into account, as well as considering the way in which these institutions co-operate with each other. The objective of breaking down silos should drive this adaptation. In line with existing OECD recommendations, steps should be taken to strengthen co-operation across policy-making departments and regulatory agencies, as well as between national and sub-national levels of government. It may also be valuable to explore how regulatory oversight bodies can help to ensure that precautionary approaches are holistic and well-balanced.

Without doubt, ensuring the systematic consideration of risk-risk trade-offs through more holistic and inclusive approaches involves rethinking several strategic policy choices. It also requires strong leadership and high-level political endorsement. These efforts will, however, remain relatively ineffective unless they trickle down — in a coherent fashion — to operational, day-to-day regulatory decisions. Moreover, approaching risk and precaution holistically requires considering not only the potential negative consequences of a given activity or innovation, but also its potential benefits, including any contribution it makes to key public policy objectives such as reducing carbon emissions and combatting climate change.

Much of the existing research and analysis regarding the PP revolves around the level of available scientific evidence and assessment, as well as related uncertainty regarding hazards and potential risks. Although frameworks and guidelines for using the PP have been proposed (e.g. by compiling lists of common circumstances when precaution might be warranted; scenario-based guidance or advice with regard to emerging risks (European Commission, 2017[7]), application remains inconsistent.

One key factor behind this inconsistency stems from the fact that socio-political and psychological elements play a determining role when it comes to precautionary decision-making. Those elements should thus be focused on as a priority. Doing so notably involves analysing factors such as the political economy and determinants of perceptions of, and responses to risk including risk perception studies. The analysis should extend tο understanding what governments and regulators can do in order to facilitate successful outcomes.

Understanding the importance of the above-mentioned factors and their interplay also requires acknowledging that the PP is not meant to instantly confer “objectivity” to the decision-making process. Nor does it have the ability to trigger some sort of “automated” decision. Rather, it is a guiding principle that needs to be interpreted and applied in a specific context, under specific constraints and — to the extent possible — be based on evidence and careful judgment. Preventing the PP from being construed as a rigid or strictly prescriptive normative blueprint is all the more necessary given that the impossibility of achieving an “optimal” precision of rules has been convincingly demonstrated (as has the unavoidable distance between rules, compliance and results). Moreover, this impossibility of “optimal” rules means that risk-based, accountable and professionally-grounded discretion at the regulatory delivery and enforcement stage is essential if the desired regulatory objectives are to be met (OECD, 2021[8]) (Diver, 1983[9]). In the case of the energy transition, interpreting and applying the PP constructively and appropriately means comparing and prioritising regulatory options in a context of multiple trade-offs and uncertainty.

A key additional aspect in this context relates to the need for enhancing inclusion and promoting reasoned transparency through ongoing public engagement and participation — as well as the promotion of extensive public deliberation about the possible hazards and risk governance (Bellaby, P. and Clark, A., 2016[10]). Governments and regulators should attempt to further articulate the notion of “concern assessment” in their risk-related regulatory practice by reviewing existing processes from that perspective. This is in addition to fully implementing existing normative guidance on stakeholder engagement in regulatory policy and governance (e.g. principle No. 2 of the OECD 2012 Recommendation of the Council on Regulatory Policy and Governance), and drawing on evidence and tools from risk communication science.

This involves adopting a proactive, duty-based approach to enhance the level of understanding of hazards and risks on a wider population basis, while mitigating the risks of manipulation of public perception (e.g. by vested interests), including through online social networks. It also involves careful monitoring and analysis of any shifts in political salience. This can necessitate further action at a political level to counter what may seem to be misinformation or fearmongering.

The state of knowledge may change over time, which in turn implies that regulations may have a need for mechanisms for revisions and adjustments.

There is a general understanding that measures based on the precautionary principle are, in principle, “provisional” in nature and thus likely to be subject to changes over time (Tosun, 2013[11]). Therefore, these measures should be reviewed when new evidence becomes available in order to assess whether precautionary action has produced the intended consequences (Garnett, K. and Parsons, D. J., 2017[12]). To that end, risk analysis as a scientific field (also referred to as risk science) should be strengthened and systematically relied upon.

In practice, evidence-based reassessment and periodic review are however not commonplace. As illustrated by the GMO example (see Chapter 3), bans tend to have a negative impact on public perception, thereby consolidating beliefs that a given technology is intolerably dangerous and should be prevented from deploying. To counter this banning-inertia and favour adaptive learning (in line with the 2021 OECD Recommendation for Agile Regulatory Governance to Harness Innovation (OECD, 2021[13]), it would be beneficial to promote the development of carefully monitored regulatory experimentation, testing, and trialling in areas with high potential to help bring about a sustainable energy transition. Examples of this include regulatory sandboxes, testbeds, innovation spaces and laboratories. Such development could learn from the example of initial pilot approvals of drugs with significant therapeutical benefits that are combined with strict supervision and pharmacovigilance systems. This approach helps to source the necessary information to progressively adapt and refine associated guidelines for using those drugs, as well as to restrict their use or ban them altogether further down the line.

While it is very difficult to know for sure whether precautionary measures are effective (Wiener and Rogers, 2002[14]), and counterfactual analysis is often not possible, adaptive learning should be actively promoted (except when global catastrophic or existential risks are at play). Measures to do so notably include carrying out horizontal comparisons across jurisdictions and designing regulatory systems that are conceived to adapt and learn from experimentation. More generally, risk management systems should be geared towards drawing the appropriate lessons from implementation of the PP and enabling informed revision on a systematic basis. This is essential to fight institutional inertia and ensure that organisational efforts and management priorities are driven by likely major hazards.

Hazard analysis should be dealt with at the level of the relevant bodies (e.g. agency) — as opposed to being left to individual inspectors (whose expert input to a wider process should however be encouraged, not least to avoid group think or leader-only-wisdom phenomena). To that end, it may be valuable to explore the potential for refocusing the mandate of those bodies towards outcomes (not simply enforcing the letter of the law), as well as to take account of situations in which hazard analysis may override risk analysis. To do so, it would be beneficial to extract relevant learnings from regulators and government agencies with experience of implementing outcome-based approaches; e.g. US Coast Guard (1996-2008), SEPA (2003-2012) and US EPA (1992-2012) – all the while bearing in mind that idiosyncratic factors may render specific approaches hard to replicate in a straightforward fashion. Relevant mandates should stress the importance of a regular review of hazards at a strategic level – the PP being applicable in cases of new emerging hazards.

Moreover, the independent assessment and review of the hazard and risk landscape in the relevant policy or regulatory area (e.g. environment, human health) should be clearly identified as an organisational duty upheld by their leadership and reflected in management practices and structure. If necessary, organisational requirements may be defined for systematic reviews to determine whether the current approach to risk is relevant and coherent. Evaluation of policy performance over time (with comparisons to both alternative policy designs and the counterfactual scenario (Wiener, 2016[15]) should also be encouraged from a whole-of-government perspective. The same goes for strengthening capacities for anticipatory analysis (e.g. foresight, horizon scanning) and embedding these activities systematically in risk governance frameworks.

A specific case concerns disasters that occur infrequently and for which it may not be possible to “judge whether the regulatory agency has struck the optimal balance in its risk management strategies”. Under these circumstances, it may be useful to explore alternative hypotheses (problems may have other causes than a lack of adequate regulatory oversight), and examine the regulatory mechanisms employed in other similarly situated regulatory environments or time periods (Carrigan and Coglianese, 2012[16]).

Independent assessment and review may usefully draw on peer review processes across government, as well as on the involvement of existing networks. Such a collaboration model has already shown its effectiveness on several occasions. This includes the joint efforts by environment protection, anti-money laundering actors and justice bodies in various environmental crime contexts, e.g. INTERPOL and EUROPOL’s joint initiatives with EPAs, police authorities and prosecutors on the prevention of waste and pollution crimes, as well as wildlife trafficking.

While politically sensitive, systematic assessment and review may also be capitalised upon to assess whether precautionary approaches and measures in place need to be modified. Such a process could consider new information or knowledge that may, for example, reduce the degree of scientific uncertainty. It may also facilitate the identification of opportunities to simplify how risks are operationally managed. Quality guidance is also necessary to that end; all the more since it has been pointed out that there is often little or no guidance as to what conditions justify a re-examination of the potential risk, and who would be responsible for producing the evidence required for risk assessment (Garnett, K. and Parsons, D. J., 2017[12]).

Adaptive learning will effectively hinge on the existence of appropriate and well-structured approaches to (risk-related) data monitoring and knowledge management. If appropriate, it may be useful to set internal organisational requirements to gather, hold, review and use relevant data and information. These requirements may be coupled with staff selection, training and incentives, and adequate reskilling/talent management measures. Repositories of both public and neutral sources of data and information may also be particularly instrumental in that respect.

Based on the analysis and examples presented in this report, there is a case for reconsidering some of the commonly accepted characterisations of precaution and precautionary measures. In a context of complex decision-making systems ridden with uncertainty, political economy struggles and multiple trade-offs across risk and policy objectives, the “default setting” matters hugely. Given major impending challenges related to climate change and sustainability more generally, regulatory choices that lock in the status quo may not be precautionary at all. In a similar vein, it is known that regulatory bans tend to render subsequent authorisation difficult due to their effect on risk perception. This warrants the consideration of alternative, non-binary options that allow for more agile and adaptable regulatory governance, at least for those risks that do not warrant a regulatory ban.

Moreover, to be prudent, regulatory decisions must not only consider the harms (e.g. potential magnitude and level of certainty), but also the potential benefits stemming from the deployment of a given technology or innovation. In other words, decision-makers need to be mindful of the (evolving) balance of risks between acting and not acting, allowing and not allowing. This holistic view may also be useful to inform differential regulatory treatment. For example, bans may be the right precautionary approach for technologies with significant downside risks, but no major expected benefits; however, the approach to managing trade-offs is likely to be more problematic for those technologies that can also bring major benefits.

Finally, a default setting of banning the development of potentially beneficial innovations also incurs the opportunity cost of not being able to gather evidence and enable an improved understanding of that technology’s real-life behaviour and impact. Controlled testing, piloting and experimentation thus appear as useful regulatory tools which — when combined with appropriate oversight and robust data strategies — can contribute to delivering superior outcomes. One such outcome is a sustainable energy transition.

  • Chapter 2 of the report provides an overview of the core elements of the PP as well as its evolution over time (including the surrounding public controversies). In addition to a general overview, the chapter includes specific examples of the PP as it relates to energy, including cases where precaution was not applied, and the lessons learned from these.

  • Chapter 3 focuses on the various interpretations of the PP as a regulatory approach, especially regarding the articulation between risk, uncertainty and precaution, as well as the methods and criteria developed to help determine whether application of the PP is warranted. It also delves into the role of public perception, human psychology, and incentive structures in the use of the PP, as well as their implications for public trust. Energy examples are specifically discussed, as public perceptions and trust issues are particularly salient in several major cases, with strong divergence between scientific risk-assessment and the generally perceived threat and harm.

  • Chapter 4 elaborates on the articulation between precautionary approaches and innovation by outlining key arguments in this long-standing debate. It showcases examples of frameworks aimed at reconciling both objectives to the extent possible.

  • Annex A presents selected examples of PP integration into national and international law.

  • Annex B presents selected cases involving the adoption of precautionary approaches.

Selected assessments and estimates in the context of precaution and risk-based regulation are provided in Chapters 2 and 3, and Annex B.


[10] Bellaby, P. and Clark, A. (2016), Might More Harm Be Done Than Good When Scientists and Engineers Engage with the Public About New Technology Before it is Fully Developed? The Case of Hydrogen Energy. International Journal of Science Education, Part B 6:, pp. 283-302.

[5] Blanc, F. et al. (2015), Understanding and addressing the Risk Regulation Reflex. Prepared for the Dutch Risk and Responsibility Programme.

[16] Carrigan, C. and C. Coglianese (2012), Oversight in Hindsight: Assessing the U.S. Regulatory System In the Wake of Calamity, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2186529.

[1] Coglianese, C. (ed.) (2016), Regulatory Excellence and lucidity, Brookings Institution Press.

[9] Diver, C. (1983), “The Optimal Precision of Administrative Rules”, The Yale Law Journal, Vol. 93/1, p. 65, https://doi.org/10.2307/796245.

[7] European Commission (2017), Future brief: The precautionary principle: decision-making under uncertainty.

[12] Garnett, K. and Parsons, D. J. (2017), Multi-Case Review of the Application of the PrecautionaryPrinciple in European Union Law and Case Law, https://onlinelibrary.wiley.com/doi/epdf/10.1111/risa.12633.

[6] Graham, J. and Wiener, J. (1995), Risk vs. Risk, Harvard University Press.

[4] Hydrogen Safety Innovation Programme (2020), Permitting process on Hydrogen Refuelling Stations. Summary of the practical guide for operators and local residents, https://opwegmetwaterstof.nl/wp-content/uploads/2020/06/Summary_Guide-permitting_process_hydrogen_refuelling_stations.pdf.

[3] Institute for Safety (2021), Hydrogen cars in parking garages - Part 1, https://nipv.nl/wp-content/uploads/2022/06/20210209-IFV-Hydrogen-cars-in-parking-garages.pdf.

[8] OECD (2021), OECD Regulatory Policy Outlook 2021, OECD Publishing, Paris, https://doi.org/10.1787/38b0fdb1-en.

[13] OECD (2021), Recommendation of the Council for Agile Regulatory Governance to Harness Innovation. OECD/LEGAL/0464.

[11] Tosun, J. (2013), Risk Regulation in Europe. Assessing the Application of the Precautionary Principle, Springer, https://link.springer.com/book/10.1007/978-1-4614-1984-6.

[2] Wiener, J. (2020), Learning to Manage the Multi-Risk World, https://doi.org/10.1111/risa.13629.

[15] Wiener, J. (2016), Precaution and Climate Change, Oxford Univ. Press, https://global.oup.com/academic/product/the-oxford-handbook-of-international-climate-change-law-9780199684601.

[14] Wiener, J. and M. Rogers (2002), Comparing precaution in the United States and Europe, pp. 317-349.


← 1. For the sake of clarity and consistency, the term precautionary principle is used in this report, though it is understood that there is no unitary definition, and precautionary approach or approaches are also commonly used terms.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2023

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.